[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7BNZX6TH94PDSNJPKKFS6H5":3,"skill-same:01M7BNZX6TH94PDSNJPKKFS6H5":48},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":25,"commitSha":26,"skillDirectory":27,"forksCount":28,"repositoryHtmlUrl":29,"repositoryHomepageUrl":30,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M7BNZX6TH94PDSNJPKKFS6H5","agentic-jujutsu",null,"Quantum-resistant, self-learning version control for AI agents with ReasoningBank intelligence and multi-agent coordination","YYC-Cube\u002FYYC3-Knowledge-Base","YYC-Cube","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F235006543?v=4","01M79J2PH4SWD2X0825X6PN748","yyc3","01M79J2PHVK3F6ZBYT8K2SCZCW",1,"MIT","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Agent-智能体\u002FYYC3-Agent-智能体-ruflo\u002Fv2\u002F.claude\u002Fskills\u002Fagentic-jujutsu\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: agentic-jujutsu\nversion: 2.3.2\ndescription: Quantum-resistant, self-learning version control for AI agents with ReasoningBank intelligence and multi-agent coordination\n---\n\n# Agentic Jujutsu - AI Agent Version Control\n\n> Quantum-ready, self-learning version control designed for multiple AI agents working simultaneously without conflicts.\n\n## When to Use This Skill\n\nUse **agentic-jujutsu** when you need:\n- ✅ Multiple AI agents modifying code simultaneously\n- ✅ Lock-free version control (23x faster than Git)\n- ✅ Self-learning AI that improves from experience\n- ✅ Quantum-resistant security for future-proof protection\n- ✅ Automatic conflict resolution (87% success rate)\n- ✅ Pattern recognition and intelligent suggestions\n- ✅ Multi-agent coordination without blocking\n\n## Quick Start\n\n### Installation\n\n```bash\nnpx agentic-jujutsu\n```\n\n### Basic Usage\n\n```javascript\nconst { JjWrapper } = require('agentic-jujutsu');\n\nconst jj = new JjWrapper();\n\n\u002F\u002F Basic operations\nawait jj.status();\nawait jj.newCommit('Add feature');\nawait jj.log(10);\n\n\u002F\u002F Self-learning trajectory\nconst id = jj.startTrajectory('Implement authentication');\nawait jj.branchCreate('feature\u002Fauth');\nawait jj.newCommit('Add auth');\njj.addToTrajectory();\njj.finalizeTrajectory(0.9, 'Clean implementation');\n\n\u002F\u002F Get AI suggestions\nconst suggestion = JSON.parse(jj.getSuggestion('Add logout feature'));\nconsole.log(`Confidence: ${suggestion.confidence}`);\n```\n\n## Core Capabilities\n\n### 1. Self-Learning with ReasoningBank\n\nTrack operations, learn patterns, and get intelligent suggestions:\n\n```javascript\n\u002F\u002F Start learning trajectory\nconst trajectoryId = jj.startTrajectory('Deploy to production');\n\n\u002F\u002F Perform operations (automatically tracked)\nawait jj.execute(['git', 'push', 'origin', 'main']);\nawait jj.branchCreate('release\u002Fv1.0');\nawait jj.newCommit('Release v1.0');\n\n\u002F\u002F Record operations to trajectory\njj.addToTrajectory();\n\n\u002F\u002F Finalize with success score (0.0-1.0) and critique\njj.finalizeTrajectory(0.95, 'Deployment successful, no issues');\n\n\u002F\u002F Later: Get AI-powered suggestions for similar tasks\nconst suggestion = JSON.parse(jj.getSuggestion('Deploy to staging'));\nconsole.log('AI Recommendation:', suggestion.reasoning);\nconsole.log('Confidence:', (suggestion.confidence * 100).toFixed(1) + '%');\nconsole.log('Expected Success:', (suggestion.expectedSuccessRate * 100).toFixed(1) + '%');\n```\n\n**Validation (v2.3.1)**:\n- ✅ Tasks must be non-empty (max 10KB)\n- ✅ Success scores must be 0.0-1.0\n- ✅ Must have operations before finalizing\n- ✅ Contexts cannot be empty\n\n### 2. Pattern Discovery\n\nAutomatically identify successful operation sequences:\n\n```javascript\n\u002F\u002F Get discovered patterns\nconst patterns = JSON.parse(jj.getPatterns());\n\npatterns.forEach(pattern => {\n    console.log(`Pattern: ${pattern.name}`);\n    console.log(`  Success rate: ${(pattern.successRate * 100).toFixed(1)}%`);\n    console.log(`  Used ${pattern.observationCount} times`);\n    console.log(`  Operations: ${pattern.operationSequence.join(' → ')}`);\n    console.log(`  Confidence: ${(pattern.confidence * 100).toFixed(1)}%`);\n});\n```\n\n### 3. Learning Statistics\n\nTrack improvement over time:\n\n```javascript\nconst stats = JSON.parse(jj.getLearningStats());\n\nconsole.log('Learning Progress:');\nconsole.log(`  Total trajectories: ${stats.totalTrajectories}`);\nconsole.log(`  Patterns discovered: ${stats.totalPatterns}`);\nconsole.log(`  Average success: ${(stats.avgSuccessRate * 100).toFixed(1)}%`);\nconsole.log(`  Improvement rate: ${(stats.improvementRate * 100).toFixed(1)}%`);\nconsole.log(`  Prediction accuracy: ${(stats.predictionAccuracy * 100).toFixed(1)}%`);\n```\n\n### 4. Multi-Agent Coordination\n\nMultiple agents work concurrently without conflicts:\n\n```javascript\n\u002F\u002F Agent 1: Developer\nconst dev = new JjWrapper();\ndev.startTrajectory('Implement feature');\nawait dev.newCommit('Add feature X');\ndev.addToTrajectory();\ndev.finalizeTrajectory(0.85);\n\n\u002F\u002F Agent 2: Reviewer (learns from Agent 1)\nconst reviewer = new JjWrapper();\nconst suggestion = JSON.parse(reviewer.getSuggestion('Review feature X'));\n\nif (suggestion.confidence > 0.7) {\n    console.log('High confidence approach:', suggestion.reasoning);\n}\n\n\u002F\u002F Agent 3: Tester (benefits from both)\nconst tester = new JjWrapper();\nconst similar = JSON.parse(tester.queryTrajectories('test feature', 5));\nconsole.log(`Found ${similar.length} similar test approaches`);\n```\n\n### 5. Quantum-Resistant Security (v2.3.0+)\n\nFast integrity verification with quantum-resistant cryptography:\n\n```javascript\nconst { generateQuantumFingerprint, verifyQuantumFingerprint } = require('agentic-jujutsu');\n\n\u002F\u002F Generate SHA3-512 fingerprint (NIST FIPS 202)\nconst data = Buffer.from('commit-data');\nconst fingerprint = generateQuantumFingerprint(data);\nconsole.log('Fingerprint:', fingerprint.toString('hex'));\n\n\u002F\u002F Verify integrity (\u003C1ms)\nconst isValid = verifyQuantumFingerprint(data, fingerprint);\nconsole.log('Valid:', isValid);\n\n\u002F\u002F HQC-128 encryption for trajectories\nconst crypto = require('crypto');\nconst key = crypto.randomBytes(32).toString('base64');\njj.enableEncryption(key);\n```\n\n### 6. Operation Tracking with AgentDB\n\nAutomatic tracking of all operations:\n\n```javascript\n\u002F\u002F Operations are tracked automatically\nawait jj.status();\nawait jj.newCommit('Fix bug');\nawait jj.rebase('main');\n\n\u002F\u002F Get operation statistics\nconst stats = JSON.parse(jj.getStats());\nconsole.log(`Total operations: ${stats.total_operations}`);\nconsole.log(`Success rate: ${(stats.success_rate * 100).toFixed(1)}%`);\nconsole.log(`Avg duration: ${stats.avg_duration_ms.toFixed(2)}ms`);\n\n\u002F\u002F Query recent operations\nconst ops = jj.getOperations(10);\nops.forEach(op => {\n    console.log(`${op.operationType}: ${op.command}`);\n    console.log(`  Duration: ${op.durationMs}ms, Success: ${op.success}`);\n});\n\n\u002F\u002F Get user operations (excludes snapshots)\nconst userOps = jj.getUserOperations(20);\n```\n\n## Advanced Use Cases\n\n### Use Case 1: Adaptive Workflow Optimization\n\nLearn and improve deployment workflows:\n\n```javascript\nasync function adaptiveDeployment(jj, environment) {\n    \u002F\u002F Get AI suggestion based on past deployments\n    const suggestion = JSON.parse(jj.getSuggestion(`Deploy to ${environment}`));\n    \n    console.log(`Deploying with ${(suggestion.confidence * 100).toFixed(0)}% confidence`);\n    console.log(`Expected duration: ${suggestion.estimatedDurationMs}ms`);\n    \n    \u002F\u002F Start tracking\n    jj.startTrajectory(`Deploy to ${environment}`);\n    \n    \u002F\u002F Execute recommended operations\n    for (const op of suggestion.recommendedOperations) {\n        console.log(`Executing: ${op}`);\n        await executeOperation(op);\n    }\n    \n    jj.addToTrajectory();\n    \n    \u002F\u002F Record outcome\n    const success = await verifyDeployment();\n    jj.finalizeTrajectory(\n        success ? 0.95 : 0.5,\n        success ? 'Deployment successful' : 'Issues detected'\n    );\n}\n```\n\n### Use Case 2: Multi-Agent Code Review\n\nCoordinate review across multiple agents:\n\n```javascript\nasync function coordinatedReview(agents) {\n    const reviews = await Promise.all(agents.map(async (agent) => {\n        const jj = new JjWrapper();\n        \n        \u002F\u002F Start review trajectory\n        jj.startTrajectory(`Review by ${agent.name}`);\n        \n        \u002F\u002F Get AI suggestion for review approach\n        const suggestion = JSON.parse(jj.getSuggestion('Code review'));\n        \n        \u002F\u002F Perform review\n        const diff = await jj.diff('@', '@-');\n        const issues = await agent.analyze(diff);\n        \n        jj.addToTrajectory();\n        jj.finalizeTrajectory(\n            issues.length === 0 ? 0.9 : 0.6,\n            `Found ${issues.length} issues`\n        );\n        \n        return { agent: agent.name, issues, suggestion };\n    }));\n    \n    \u002F\u002F Aggregate learning from all agents\n    return reviews;\n}\n```\n\n### Use Case 3: Error Pattern Detection\n\nLearn from failures to prevent future issues:\n\n```javascript\nasync function smartMerge(jj, branch) {\n    \u002F\u002F Query similar merge attempts\n    const similar = JSON.parse(jj.queryTrajectories(`merge ${branch}`, 10));\n    \n    \u002F\u002F Analyze past failures\n    const failures = similar.filter(t => t.successScore \u003C 0.5);\n    \n    if (failures.length > 0) {\n        console.log('⚠️ Similar merges failed in the past:');\n        failures.forEach(f => {\n            if (f.critique) {\n                console.log(`  - ${f.critique}`);\n            }\n        });\n    }\n    \n    \u002F\u002F Get AI recommendation\n    const suggestion = JSON.parse(jj.getSuggestion(`merge ${branch}`));\n    \n    if (suggestion.confidence \u003C 0.7) {\n        console.log('⚠️ Low confidence. Recommended steps:');\n        suggestion.recommendedOperations.forEach(op => console.log(`  - ${op}`));\n    }\n    \n    \u002F\u002F Execute merge with tracking\n    jj.startTrajectory(`Merge ${branch}`);\n    try {\n        await jj.execute(['merge', branch]);\n        jj.addToTrajectory();\n        jj.finalizeTrajectory(0.9, 'Merge successful');\n    } catch (err) {\n        jj.addToTrajectory();\n        jj.finalizeTrajectory(0.3, `Merge failed: ${err.message}`);\n        throw err;\n    }\n}\n```\n\n### Use Case 4: Continuous Learning Loop\n\nImplement a self-improving agent:\n\n```javascript\nclass SelfImprovingAgent {\n    constructor() {\n        this.jj = new JjWrapper();\n    }\n    \n    async performTask(taskDescription) {\n        \u002F\u002F Get AI suggestion\n        const suggestion = JSON.parse(this.jj.getSuggestion(taskDescription));\n        \n        console.log(`Task: ${taskDescription}`);\n        console.log(`AI Confidence: ${(suggestion.confidence * 100).toFixed(1)}%`);\n        console.log(`Expected Success: ${(suggestion.expectedSuccessRate * 100).toFixed(1)}%`);\n        \n        \u002F\u002F Start trajectory\n        this.jj.startTrajectory(taskDescription);\n        \n        \u002F\u002F Execute with recommended approach\n        const startTime = Date.now();\n        let success = false;\n        \n        try {\n            for (const op of suggestion.recommendedOperations) {\n                await this.execute(op);\n            }\n            success = true;\n        } catch (err) {\n            console.error('Task failed:', err.message);\n        }\n        \n        const duration = Date.now() - startTime;\n        \n        \u002F\u002F Record learning\n        this.jj.addToTrajectory();\n        this.jj.finalizeTrajectory(\n            success ? 0.9 : 0.4,\n            success \n                ? `Completed in ${duration}ms using ${suggestion.recommendedOperations.length} operations`\n                : `Failed after ${duration}ms`\n        );\n        \n        \u002F\u002F Check improvement\n        const stats = JSON.parse(this.jj.getLearningStats());\n        console.log(`Improvement rate: ${(stats.improvementRate * 100).toFixed(1)}%`);\n        \n        return success;\n    }\n    \n    async execute(operation) {\n        \u002F\u002F Execute operation logic\n    }\n}\n\n\u002F\u002F Usage\nconst agent = new SelfImprovingAgent();\n\n\u002F\u002F Agent improves over time\nfor (let i = 1; i \u003C= 10; i++) {\n    console.log(`\\n--- Attempt ${i} ---`);\n    await agent.performTask('Deploy application');\n}\n```\n\n## API Reference\n\n### Core Methods\n\n| Method | Description | Returns |\n|--------|-------------|---------|\n| `new JjWrapper()` | Create wrapper instance | JjWrapper |\n| `status()` | Get repository status | Promise\u003CJjResult> |\n| `newCommit(msg)` | Create new commit | Promise\u003CJjResult> |\n| `log(limit)` | Show commit history | Promise\u003CJjCommit[]> |\n| `diff(from, to)` | Show differences | Promise\u003CJjDiff> |\n| `branchCreate(name, rev?)` | Create branch | Promise\u003CJjResult> |\n| `rebase(source, dest)` | Rebase commits | Promise\u003CJjResult> |\n\n### ReasoningBank Methods\n\n| Method | Description | Returns |\n|--------|-------------|---------|\n| `startTrajectory(task)` | Begin learning trajectory | string (trajectory ID) |\n| `addToTrajectory()` | Add recent operations | void |\n| `finalizeTrajectory(score, critique?)` | Complete trajectory (score: 0.0-1.0) | void |\n| `getSuggestion(task)` | Get AI recommendation | JSON: DecisionSuggestion |\n| `getLearningStats()` | Get learning metrics | JSON: LearningStats |\n| `getPatterns()` | Get discovered patterns | JSON: Pattern[] |\n| `queryTrajectories(task, limit)` | Find similar trajectories | JSON: Trajectory[] |\n| `resetLearning()` | Clear learned data | void |\n\n### AgentDB Methods\n\n| Method | Description | Returns |\n|--------|-------------|---------|\n| `getStats()` | Get operation statistics | JSON: Stats |\n| `getOperations(limit)` | Get recent operations | JjOperation[] |\n| `getUserOperations(limit)` | Get user operations only | JjOperation[] |\n| `clearLog()` | Clear operation log | void |\n\n### Quantum Security Methods (v2.3.0+)\n\n| Method | Description | Returns |\n|--------|-------------|---------|\n| `generateQuantumFingerprint(data)` | Generate SHA3-512 fingerprint | Buffer (64 bytes) |\n| `verifyQuantumFingerprint(data, fp)` | Verify fingerprint | boolean |\n| `enableEncryption(key, pubKey?)` | Enable HQC-128 encryption | void |\n| `disableEncryption()` | Disable encryption | void |\n| `isEncryptionEnabled()` | Check encryption status | boolean |\n\n## Performance Characteristics\n\n| Metric | Git | Agentic Jujutsu |\n|--------|-----|-----------------|\n| Concurrent commits | 15 ops\u002Fs | 350 ops\u002Fs (23x) |\n| Context switching | 500-1000ms | 50-100ms (10x) |\n| Conflict resolution | 30-40% auto | 87% auto (2.5x) |\n| Lock waiting | 50 min\u002Fday | 0 min (∞) |\n| Quantum fingerprints | N\u002FA | \u003C1ms |\n\n## Best Practices\n\n### 1. Trajectory Management\n\n```javascript\n\u002F\u002F ✅ Good: Meaningful task descriptions\njj.startTrajectory('Implement user authentication with JWT');\n\n\u002F\u002F ❌ Bad: Vague descriptions\njj.startTrajectory('fix stuff');\n\n\u002F\u002F ✅ Good: Honest success scores\njj.finalizeTrajectory(0.7, 'Works but needs refactoring');\n\n\u002F\u002F ❌ Bad: Always 1.0\njj.finalizeTrajectory(1.0, 'Perfect!'); \u002F\u002F Prevents learning\n```\n\n### 2. Pattern Recognition\n\n```javascript\n\u002F\u002F ✅ Good: Let patterns emerge naturally\nfor (let i = 0; i \u003C 10; i++) {\n    jj.startTrajectory('Deploy feature');\n    await deploy();\n    jj.addToTrajectory();\n    jj.finalizeTrajectory(wasSuccessful ? 0.9 : 0.5);\n}\n\n\u002F\u002F ❌ Bad: Not recording outcomes\nawait deploy(); \u002F\u002F No learning\n```\n\n### 3. Multi-Agent Coordination\n\n```javascript\n\u002F\u002F ✅ Good: Concurrent operations\nconst agents = ['agent1', 'agent2', 'agent3'];\nawait Promise.all(agents.map(async (agent) => {\n    const jj = new JjWrapper();\n    \u002F\u002F Each agent works independently\n    await jj.newCommit(`Changes by ${agent}`);\n}));\n\n\u002F\u002F ❌ Bad: Sequential with locks\nfor (const agent of agents) {\n    await agent.waitForLock(); \u002F\u002F Not needed!\n    await agent.commit();\n}\n```\n\n### 4. Error Handling\n\n```javascript\n\u002F\u002F ✅ Good: Record failures with details\ntry {\n    await jj.execute(['complex-operation']);\n    jj.finalizeTrajectory(0.9);\n} catch (err) {\n    jj.finalizeTrajectory(0.3, `Failed: ${err.message}. Root cause: ...`);\n}\n\n\u002F\u002F ❌ Bad: Silent failures\ntry {\n    await jj.execute(['operation']);\n} catch (err) {\n    \u002F\u002F No learning from failure\n}\n```\n\n## Validation Rules (v2.3.1+)\n\n### Task Description\n- ✅ Cannot be empty or whitespace-only\n- ✅ Maximum length: 10,000 bytes\n- ✅ Automatically trimmed\n\n### Success Score\n- ✅ Must be finite (not NaN or Infinity)\n- ✅ Must be between 0.0 and 1.0 (inclusive)\n\n### Operations\n- ✅ Must have at least one operation before finalizing\n\n### Context\n- ✅ Cannot be empty\n- ✅ Keys cannot be empty or whitespace-only\n- ✅ Keys max 1,000 bytes, values max 10,000 bytes\n\n## Troubleshooting\n\n### Issue: Low Confidence Suggestions\n\n```javascript\nconst suggestion = JSON.parse(jj.getSuggestion('new task'));\n\nif (suggestion.confidence \u003C 0.5) {\n    \u002F\u002F Not enough data - check learning stats\n    const stats = JSON.parse(jj.getLearningStats());\n    console.log(`Need more data. Current trajectories: ${stats.totalTrajectories}`);\n    \n    \u002F\u002F Recommend: Record 5-10 trajectories first\n}\n```\n\n### Issue: Validation Errors\n\n```javascript\ntry {\n    jj.startTrajectory(''); \u002F\u002F Empty task\n} catch (err) {\n    if (err.message.includes('Validation error')) {\n        console.log('Invalid input:', err.message);\n        \u002F\u002F Use non-empty, meaningful task description\n    }\n}\n\ntry {\n    jj.finalizeTrajectory(1.5); \u002F\u002F Score > 1.0\n} catch (err) {\n    \u002F\u002F Use score between 0.0 and 1.0\n    jj.finalizeTrajectory(Math.max(0, Math.min(1, score)));\n}\n```\n\n### Issue: No Patterns Discovered\n\n```javascript\nconst patterns = JSON.parse(jj.getPatterns());\n\nif (patterns.length === 0) {\n    \u002F\u002F Need more trajectories with >70% success\n    \u002F\u002F Record at least 3-5 successful trajectories\n}\n```\n\n## Examples\n\n### Example 1: Simple Learning Workflow\n\n```javascript\nconst { JjWrapper } = require('agentic-jujutsu');\n\nasync function learnFromWork() {\n    const jj = new JjWrapper();\n    \n    \u002F\u002F Start tracking\n    jj.startTrajectory('Add user profile feature');\n    \n    \u002F\u002F Do work\n    await jj.branchCreate('feature\u002Fuser-profile');\n    await jj.newCommit('Add user profile model');\n    await jj.newCommit('Add profile API endpoints');\n    await jj.newCommit('Add profile UI');\n    \n    \u002F\u002F Record operations\n    jj.addToTrajectory();\n    \n    \u002F\u002F Finalize with result\n    jj.finalizeTrajectory(0.85, 'Feature complete, minor styling issues remain');\n    \n    \u002F\u002F Next time, get suggestions\n    const suggestion = JSON.parse(jj.getSuggestion('Add settings page'));\n    console.log('AI suggests:', suggestion.reasoning);\n}\n```\n\n### Example 2: Multi-Agent Swarm\n\n```javascript\nasync function agentSwarm(taskList) {\n    const agents = taskList.map((task, i) => ({\n        name: `agent-${i}`,\n        jj: new JjWrapper(),\n        task\n    }));\n    \n    \u002F\u002F All agents work concurrently (no conflicts!)\n    const results = await Promise.all(agents.map(async (agent) => {\n        agent.jj.startTrajectory(agent.task);\n        \n        \u002F\u002F Get AI suggestion\n        const suggestion = JSON.parse(agent.jj.getSuggestion(agent.task));\n        \n        \u002F\u002F Execute task\n        const success = await executeTask(agent, suggestion);\n        \n        agent.jj.addToTrajectory();\n        agent.jj.finalizeTrajectory(success ? 0.9 : 0.5);\n        \n        return { agent: agent.name, success };\n    }));\n    \n    console.log('Results:', results);\n}\n```\n\n## Related Documentation\n\n- **NPM Package**: https:\u002F\u002Fnpmjs.com\u002Fpackage\u002Fagentic-jujutsu\n- **GitHub**: https:\u002F\u002Fgithub.com\u002Fruvnet\u002Fagentic-flow\u002Ftree\u002Fmain\u002Fpackages\u002Fagentic-jujutsu\n- **Full README**: See package README.md\n- **Validation Guide**: docs\u002FVALIDATION_FIXES_v2.3.1.md\n- **AgentDB Guide**: docs\u002FAGENTDB_GUIDE.md\n\n## Version History\n\n- **v2.3.2** - Documentation updates\n- **v2.3.1** - Validation fixes for ReasoningBank\n- **v2.3.0** - Quantum-resistant security with @qudag\u002Fnapi-core\n- **v2.1.0** - Self-learning AI with ReasoningBank\n- **v2.0.0** - Zero-dependency installation with embedded jj binary\n\n---\n\n**Status**: ✅ Production Ready\n**License**: MIT\n**Maintained**: Active\n","# Agentic Jujutsu - AI Agent Version Control\n\n> Quantum-ready, self-learning version control designed for multiple AI agents working simultaneously without conflicts.\n\n## When to Use This Skill\n\nUse **agentic-jujutsu** when you need:\n- ✅ Multiple AI agents modifying code simultaneously\n- ✅ Lock-free version control (23x faster than Git)\n- ✅ Self-learning AI that improves from experience\n- ✅ Quantum-resistant security for future-proof protection\n- ✅ Automatic conflict resolution (87% success rate)\n- ✅ Pattern recognition and intelligent suggestions\n- ✅ Multi-agent coordination without blocking\n\n## Quick Start\n\n### Installation\n\n```bash\nnpx agentic-jujutsu\n```\n\n### Basic Usage\n\n```javascript\nconst { JjWrapper } = require('agentic-jujutsu');\n\nconst jj = new JjWrapper();\n\n\u002F\u002F Basic operations\nawait jj.status();\nawait jj.newCommit('Add feature');\nawait jj.log(10);\n\n\u002F\u002F Self-learning trajectory\nconst id = jj.startTrajectory('Implement authentication');\nawait jj.branchCreate('feature\u002Fauth');\nawait jj.newCommit('Add auth');\njj.addToTrajectory();\njj.finalizeTrajectory(0.9, 'Clean implementation');\n\n\u002F\u002F Get AI suggestions\nconst suggestion = JSON.parse(jj.getSuggestion('Add logout feature'));\nconsole.log(`Confidence: ${suggestion.confidence}`);\n```\n\n## Core Capabilities\n\n### 1. Self-Learning with ReasoningBank\n\nTrack operations, learn patterns, and get intelligent suggestions:\n\n```javascript\n\u002F\u002F Start learning trajectory\nconst trajectoryId = jj.startTrajectory('Deploy to production');\n\n\u002F\u002F Perform operations (automatically tracked)\nawait jj.execute(['git', 'push', 'origin', 'main']);\nawait jj.branchCreate('release\u002Fv1.0');\nawait jj.newCommit('Release v1.0');\n\n\u002F\u002F Record operations to trajectory\njj.addToTrajectory();\n\n\u002F\u002F Finalize with success score (0.0-1.0) and critique\njj.finalizeTrajectory(0.95, 'Deployment successful, no issues');\n\n\u002F\u002F Later: Get AI-powered suggestions for similar tasks\nconst suggestion = JSON.parse(jj.getSuggestion('Deploy to staging'));\nconsole.log('AI Recommendation:', suggestion.reasoning);\nconsole.log('Confidence:', (suggestion.confidence * 100).toFixed(1) + '%');\nconsole.log('Expected Success:', (suggestion.expectedSuccessRate * 100).toFixed(1) + '%');\n```\n\n**Validation (v2.3.1)**:\n- ✅ Tasks must be non-empty (max 10KB)\n- ✅ Success scores must be 0.0-1.0\n- ✅ Must have operations before finalizing\n- ✅ Contexts cannot be empty\n\n### 2. Pattern Discovery\n\nAutomatically identify successful operation sequences:\n\n```javascript\n\u002F\u002F Get discovered patterns\nconst patterns = JSON.parse(jj.getPatterns());\n\npatterns.forEach(pattern => {\n    console.log(`Pattern: ${pattern.name}`);\n    console.log(`  Success rate: ${(pattern.successRate * 100).toFixed(1)}%`);\n    console.log(`  Used ${pattern.observationCount} times`);\n    console.log(`  Operations: ${pattern.operationSequence.join(' → ')}`);\n    console.log(`  Confidence: ${(pattern.confidence * 100).toFixed(1)}%`);\n});\n```\n\n### 3. Learning Statistics\n\nTrack improvement over time:\n\n```javascript\nconst stats = JSON.parse(jj.getLearningStats());\n\nconsole.log('Learning Progress:');\nconsole.log(`  Total trajectories: ${stats.totalTrajectories}`);\nconsole.log(`  Patterns discovered: ${stats.totalPatterns}`);\nconsole.log(`  Average success: ${(stats.avgSuccessRate * 100).toFixed(1)}%`);\nconsole.log(`  Improvement rate: ${(stats.improvementRate * 100).toFixed(1)}%`);\nconsole.log(`  Prediction accuracy: ${(stats.predictionAccuracy * 100).toFixed(1)}%`);\n```\n\n### 4. Multi-Agent Coordination\n\nMultiple agents work concurrently without conflicts:\n\n```javascript\n\u002F\u002F Agent 1: Developer\nconst dev = new JjWrapper();\ndev.startTrajectory('Implement feature');\nawait dev.newCommit('Add feature X');\ndev.addToTrajectory();\ndev.finalizeTrajectory(0.85);\n\n\u002F\u002F Agent 2: Reviewer (learns from Agent 1)\nconst reviewer = new JjWrapper();\nconst suggestion = JSON.parse(reviewer.getSuggestion('Review feature X'));\n\nif (suggestion.confidence > 0.7) {\n    console.log('High confidence approach:', suggestion.reasoning);\n}\n\n\u002F\u002F Agent 3: Tester (benefits from both)\nconst tester = new JjWrapper();\nconst similar = JSON.parse(tester.queryTrajectories('test feature', 5));\nconsole.log(`Found ${similar.length} similar test approaches`);\n```\n\n### 5. Quantum-Resistant Security (v2.3.0+)\n\nFast integrity verification with quantum-resistant cryptography:\n\n```javascript\nconst { generateQuantumFingerprint, verifyQuantumFingerprint } = require('agentic-jujutsu');\n\n\u002F\u002F Generate SHA3-512 fingerprint (NIST FIPS 202)\nconst data = Buffer.from('commit-data');\nconst fingerprint = generateQuantumFingerprint(data);\nconsole.log('Fingerprint:', fingerprint.toString('hex'));\n\n\u002F\u002F Verify integrity (\u003C1ms)\nconst isValid = verifyQuantumFingerprint(data, fingerprint);\nconsole.log('Valid:', isValid);\n\n\u002F\u002F HQC-128 encryption for trajectories\nconst crypto = require('crypto');\nconst key = crypto.randomBytes(32).toString('base64');\njj.enableEncryption(key);\n```\n\n### 6. Operation Tracking with AgentDB\n\nAutomatic tracking of all operations:\n\n```javascript\n\u002F\u002F Operations are tracked automatically\nawait jj.status();\nawait jj.newCommit('Fix bug');\nawait jj.rebase('main');\n\n\u002F\u002F Get operation statistics\nconst stats = JSON.parse(jj.getStats());\nconsole.log(`Total operations: ${stats.total_operations}`);\nconsole.log(`Success rate: ${(stats.success_rate * 100).toFixed(1)}%`);\nconsole.log(`Avg duration: ${stats.avg_duration_ms.toFixed(2)}ms`);\n\n\u002F\u002F Query recent operations\nconst ops = jj.getOperations(10);\nops.forEach(op => {\n    console.log(`${op.operationType}: ${op.command}`);\n    console.log(`  Duration: ${op.durationMs}ms, Success: ${op.success}`);\n});\n\n\u002F\u002F Get user operations (excludes snapshots)\nconst userOps = jj.getUserOperations(20);\n```\n\n## Advanced Use Cases\n\n### Use Case 1: Adaptive Workflow Optimization\n\nLearn and improve deployment workflows:\n\n```javascript\nasync function adaptiveDeployment(jj, environment) {\n    \u002F\u002F Get AI suggestion based on past deployments\n    const suggestion = JSON.parse(jj.getSuggestion(`Deploy to ${environment}`));\n    \n    console.log(`Deploying with ${(suggestion.confidence * 100).toFixed(0)}% confidence`);\n    console.log(`Expected duration: ${suggestion.estimatedDurationMs}ms`);\n    \n    \u002F\u002F Start tracking\n    jj.startTrajectory(`Deploy to ${environment}`);\n    \n    \u002F\u002F Execute recommended operations\n    for (const op of suggestion.recommendedOperations) {\n        console.log(`Executing: ${op}`);\n        await executeOperation(op);\n    }\n    \n    jj.addToTrajectory();\n    \n    \u002F\u002F Record outcome\n    const success = await verifyDeployment();\n    jj.finalizeTrajectory(\n        success ? 0.95 : 0.5,\n        success ? 'Deployment successful' : 'Issues detected'\n    );\n}\n```\n\n### Use Case 2: Multi-Agent Code Review\n\nCoordinate review across multiple agents:\n\n```javascript\nasync function coordinatedReview(agents) {\n    const reviews = await Promise.all(agents.map(async (agent) => {\n        const jj = new JjWrapper();\n        \n        \u002F\u002F Start review trajectory\n        jj.startTrajectory(`Review by ${agent.name}`);\n        \n        \u002F\u002F Get AI suggestion for review approach\n        const suggestion = JSON.parse(jj.getSuggestion('Code review'));\n        \n        \u002F\u002F Perform review\n        const diff = await jj.diff('@', '@-');\n        const issues = await agent.analyze(diff);\n        \n        jj.addToTrajectory();\n        jj.finalizeTrajectory(\n            issues.length === 0 ? 0.9 : 0.6,\n            `Found ${issues.length} issues`\n        );\n        \n        return { agent: agent.name, issues, suggestion };\n    }));\n    \n    \u002F\u002F Aggregate learning from all agents\n    return reviews;\n}\n```\n\n### Use Case 3: Error Pattern Detection\n\nLearn from failures to prevent future issues:\n\n```javascript\nasync function smartMerge(jj, branch) {\n    \u002F\u002F Query similar merge attempts\n    const similar = JSON.parse(jj.queryTrajectories(`merge ${branch}`, 10));\n    \n    \u002F\u002F Analyze past failures\n    const failures = similar.filter(t => t.successScore \u003C 0.5);\n    \n    if (failures.length > 0) {\n        console.log('⚠️ Similar merges failed in the past:');\n        failures.forEach(f => {\n            if (f.critique) {\n                console.log(`  - ${f.critique}`);\n            }\n        });\n    }\n    \n    \u002F\u002F Get AI recommendation\n    const suggestion = JSON.parse(jj.getSuggestion(`merge ${branch}`));\n    \n    if (suggestion.confidence \u003C 0.7) {\n        console.log('⚠️ Low confidence. Recommended steps:');\n        suggestion.recommendedOperations.forEach(op => console.log(`  - ${op}`));\n    }\n    \n    \u002F\u002F Execute merge with tracking\n    jj.startTrajectory(`Merge ${branch}`);\n    try {\n        await jj.execute(['merge', branch]);\n        jj.addToTrajectory();\n        jj.finalizeTrajectory(0.9, 'Merge successful');\n    } catch (err) {\n        jj.addToTrajectory();\n        jj.finalizeTrajectory(0.3, `Merge failed: ${err.message}`);\n        throw err;\n    }\n}\n```\n\n### Use Case 4: Continuous Learning Loop\n\nImplement a self-improving agent:\n\n```javascript\nclass SelfImprovingAgent {\n    constructor() {\n        this.jj = new JjWrapper();\n    }\n    \n    async performTask(taskDescription) {\n        \u002F\u002F Get AI suggestion\n        const suggestion = JSON.parse(this.jj.getSuggestion(taskDescription));\n        \n        console.log(`Task: ${taskDescription}`);\n        console.log(`AI Confidence: ${(suggestion.confidence * 100).toFixed(1)}%`);\n        console.log(`Expected Success: ${(suggestion.expectedSuccessRate * 100).toFixed(1)}%`);\n        \n        \u002F\u002F Start trajectory\n        this.jj.startTrajectory(taskDescription);\n        \n        \u002F\u002F Execute with recommended approach\n        const startTime = Date.now();\n        let success = false;\n        \n        try {\n            for (const op of suggestion.recommendedOperations) {\n                await this.execute(op);\n            }\n            success = true;\n        } catch (err) {\n            console.error('Task failed:', err.message);\n        }\n        \n        const duration = Date.now() - startTime;\n        \n        \u002F\u002F Record learning\n        this.jj.addToTrajectory();\n        this.jj.finalizeTrajectory(\n            success ? 0.9 : 0.4,\n            success \n                ? `Completed in ${duration}ms using ${suggestion.recommendedOperations.length} operations`\n                : `Failed after ${duration}ms`\n        );\n        \n        \u002F\u002F Check improvement\n        const stats = JSON.parse(this.jj.getLearningStats());\n        console.log(`Improvement rate: ${(stats.improvementRate * 100).toFixed(1)}%`);\n        \n        return success;\n    }\n    \n    async execute(operation) {\n        \u002F\u002F Execute operation logic\n    }\n}\n\n\u002F\u002F Usage\nconst agent = new SelfImprovingAgent();\n\n\u002F\u002F Agent improves over time\nfor (let i = 1; i \u003C= 10; i++) {\n    console.log(`\\n--- Attempt ${i} ---`);\n    await agent.performTask('Deploy application');\n}\n```\n\n## API Reference\n\n### Core Methods\n\n| Method | Description | Returns |\n|--------|-------------|---------|\n| `new JjWrapper()` | Create wrapper instance | JjWrapper |\n| `status()` | Get repository status | Promise\u003CJjResult> |\n| `newCommit(msg)` | Create new commit | Promise\u003CJjResult> |\n| `log(limit)` | Show commit history | Promise\u003CJjCommit[]> |\n| `diff(from, to)` | Show differences | Promise\u003CJjDiff> |\n| `branchCreate(name, rev?)` | Create branch | Promise\u003CJjResult> |\n| `rebase(source, dest)` | Rebase commits | Promise\u003CJjResult> |\n\n### ReasoningBank Methods\n\n| Method | Description | Returns |\n|--------|-------------|---------|\n| `startTrajectory(task)` | Begin learning trajectory | string (trajectory ID) |\n| `addToTrajectory()` | Add recent operations | void |\n| `finalizeTrajectory(score, critique?)` | Complete trajectory (score: 0.0-1.0) | void |\n| `getSuggestion(task)` | Get AI recommendation | JSON: DecisionSuggestion |\n| `getLearningStats()` | Get learning metrics | JSON: LearningStats |\n| `getPatterns()` | Get discovered patterns | JSON: Pattern[] |\n| `queryTrajectories(task, limit)` | Find similar trajectories | JSON: Trajectory[] |\n| `resetLearning()` | Clear learned data | void |\n\n### AgentDB Methods\n\n| Method | Description | Returns |\n|--------|-------------|---------|\n| `getStats()` | Get operation statistics | JSON: Stats |\n| `getOperations(limit)` | Get recent operations | JjOperation[] |\n| `getUserOperations(limit)` | Get user operations only | JjOperation[] |\n| `clearLog()` | Clear operation log | void |\n\n### Quantum Security Methods (v2.3.0+)\n\n| Method | Description | Returns |\n|--------|-------------|---------|\n| `generateQuantumFingerprint(data)` | Generate SHA3-512 fingerprint | Buffer (64 bytes) |\n| `verifyQuantumFingerprint(data, fp)` | Verify fingerprint | boolean |\n| `enableEncryption(key, pubKey?)` | Enable HQC-128 encryption | void |\n| `disableEncryption()` | Disable encryption | void |\n| `isEncryptionEnabled()` | Check encryption status | boolean |\n\n## Performance Characteristics\n\n| Metric | Git | Agentic Jujutsu |\n|--------|-----|-----------------|\n| Concurrent commits | 15 ops\u002Fs | 350 ops\u002Fs (23x) |\n| Context switching | 500-1000ms | 50-100ms (10x) |\n| Conflict resolution | 30-40% auto | 87% auto (2.5x) |\n| Lock waiting | 50 min\u002Fday | 0 min (∞) |\n| Quantum fingerprints | N\u002FA | \u003C1ms |\n\n## Best Practices\n\n### 1. Trajectory Management\n\n```javascript\n\u002F\u002F ✅ Good: Meaningful task descriptions\njj.startTrajectory('Implement user authentication with JWT');\n\n\u002F\u002F ❌ Bad: Vague descriptions\njj.startTrajectory('fix stuff');\n\n\u002F\u002F ✅ Good: Honest success scores\njj.finalizeTrajectory(0.7, 'Works but needs refactoring');\n\n\u002F\u002F ❌ Bad: Always 1.0\njj.finalizeTrajectory(1.0, 'Perfect!'); \u002F\u002F Prevents learning\n```\n\n### 2. Pattern Recognition\n\n```javascript\n\u002F\u002F ✅ Good: Let patterns emerge naturally\nfor (let i = 0; i \u003C 10; i++) {\n    jj.startTrajectory('Deploy feature');\n    await deploy();\n    jj.addToTrajectory();\n    jj.finalizeTrajectory(wasSuccessful ? 0.9 : 0.5);\n}\n\n\u002F\u002F ❌ Bad: Not recording outcomes\nawait deploy(); \u002F\u002F No learning\n```\n\n### 3. Multi-Agent Coordination\n\n```javascript\n\u002F\u002F ✅ Good: Concurrent operations\nconst agents = ['agent1', 'agent2', 'agent3'];\nawait Promise.all(agents.map(async (agent) => {\n    const jj = new JjWrapper();\n    \u002F\u002F Each agent works independently\n    await jj.newCommit(`Changes by ${agent}`);\n}));\n\n\u002F\u002F ❌ Bad: Sequential with locks\nfor (const agent of agents) {\n    await agent.waitForLock(); \u002F\u002F Not needed!\n    await agent.commit();\n}\n```\n\n### 4. Error Handling\n\n```javascript\n\u002F\u002F ✅ Good: Record failures with details\ntry {\n    await jj.execute(['complex-operation']);\n    jj.finalizeTrajectory(0.9);\n} catch (err) {\n    jj.finalizeTrajectory(0.3, `Failed: ${err.message}. Root cause: ...`);\n}\n\n\u002F\u002F ❌ Bad: Silent failures\ntry {\n    await jj.execute(['operation']);\n} catch (err) {\n    \u002F\u002F No learning from failure\n}\n```\n\n## Validation Rules (v2.3.1+)\n\n### Task Description\n- ✅ Cannot be empty or whitespace-only\n- ✅ Maximum length: 10,000 bytes\n- ✅ Automatically trimmed\n\n### Success Score\n- ✅ Must be finite (not NaN or Infinity)\n- ✅ Must be between 0.0 and 1.0 (inclusive)\n\n### Operations\n- ✅ Must have at least one operation before finalizing\n\n### Context\n- ✅ Cannot be empty\n- ✅ Keys cannot be empty or whitespace-only\n- ✅ Keys max 1,000 bytes, values max 10,000 bytes\n\n## Troubleshooting\n\n### Issue: Low Confidence Suggestions\n\n```javascript\nconst suggestion = JSON.parse(jj.getSuggestion('new task'));\n\nif (suggestion.confidence \u003C 0.5) {\n    \u002F\u002F Not enough data - check learning stats\n    const stats = JSON.parse(jj.getLearningStats());\n    console.log(`Need more data. Current trajectories: ${stats.totalTrajectories}`);\n    \n    \u002F\u002F Recommend: Record 5-10 trajectories first\n}\n```\n\n### Issue: Validation Errors\n\n```javascript\ntry {\n    jj.startTrajectory(''); \u002F\u002F Empty task\n} catch (err) {\n    if (err.message.includes('Validation error')) {\n        console.log('Invalid input:', err.message);\n        \u002F\u002F Use non-empty, meaningful task description\n    }\n}\n\ntry {\n    jj.finalizeTrajectory(1.5); \u002F\u002F Score > 1.0\n} catch (err) {\n    \u002F\u002F Use score between 0.0 and 1.0\n    jj.finalizeTrajectory(Math.max(0, Math.min(1, score)));\n}\n```\n\n### Issue: No Patterns Discovered\n\n```javascript\nconst patterns = JSON.parse(jj.getPatterns());\n\nif (patterns.length === 0) {\n    \u002F\u002F Need more trajectories with >70% success\n    \u002F\u002F Record at least 3-5 successful trajectories\n}\n```\n\n## Examples\n\n### Example 1: Simple Learning Workflow\n\n```javascript\nconst { JjWrapper } = require('agentic-jujutsu');\n\nasync function learnFromWork() {\n    const jj = new JjWrapper();\n    \n    \u002F\u002F Start tracking\n    jj.startTrajectory('Add user profile feature');\n    \n    \u002F\u002F Do work\n    await jj.branchCreate('feature\u002Fuser-profile');\n    await jj.newCommit('Add user profile model');\n    await jj.newCommit('Add profile API endpoints');\n    await jj.newCommit('Add profile UI');\n    \n    \u002F\u002F Record operations\n    jj.addToTrajectory();\n    \n    \u002F\u002F Finalize with result\n    jj.finalizeTrajectory(0.85, 'Feature complete, minor styling issues remain');\n    \n    \u002F\u002F Next time, get suggestions\n    const suggestion = JSON.parse(jj.getSuggestion('Add settings page'));\n    console.log('AI suggests:', suggestion.reasoning);\n}\n```\n\n### Example 2: Multi-Agent Swarm\n\n```javascript\nasync function agentSwarm(taskList) {\n    const agents = taskList.map((task, i) => ({\n        name: `agent-${i}`,\n        jj: new JjWrapper(),\n        task\n    }));\n    \n    \u002F\u002F All agents work concurrently (no conflicts!)\n    const results = await Promise.all(agents.map(async (agent) => {\n        agent.jj.startTrajectory(agent.task);\n        \n        \u002F\u002F Get AI suggestion\n        const suggestion = JSON.parse(agent.jj.getSuggestion(agent.task));\n        \n        \u002F\u002F Execute task\n        const success = await executeTask(agent, suggestion);\n        \n        agent.jj.addToTrajectory();\n        agent.jj.finalizeTrajectory(success ? 0.9 : 0.5);\n        \n        return { agent: agent.name, success };\n    }));\n    \n    console.log('Results:', results);\n}\n```\n\n## Related Documentation\n\n- **NPM Package**: https:\u002F\u002Fnpmjs.com\u002Fpackage\u002Fagentic-jujutsu\n- **GitHub**: https:\u002F\u002Fgithub.com\u002Fruvnet\u002Fagentic-flow\u002Ftree\u002Fmain\u002Fpackages\u002Fagentic-jujutsu\n- **Full README**: See package README.md\n- **Validation Guide**: docs\u002FVALIDATION_FIXES_v2.3.1.md\n- **AgentDB Guide**: docs\u002FAGENTDB_GUIDE.md\n\n## Version History\n\n- **v2.3.2** - Documentation updates\n- **v2.3.1** - Validation fixes for ReasoningBank\n- **v2.3.0** - Quantum-resistant security with @qudag\u002Fnapi-core\n- **v2.1.0** - Self-learning AI with ReasoningBank\n- **v2.0.0** - Zero-dependency installation with embedded jj binary\n\n---\n\n**Status**: ✅ Production Ready\n**License**: MIT\n**Maintained**: Active\n",{"name":5,"version":24,"description":7},"2.3.2","main","87afe6ce96220accf50a345dea46784c8c9f210e","YYC3-Agent-智能体\u002FYYC3-Agent-智能体-ruflo\u002Fv2\u002F.claude\u002Fskills\u002Fagentic-jujutsu",0,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base","https:\u002F\u002Fkb.yyc3.top\u002F","https:\u002F\u002Fgithub.com\u002FYYC-Cube","万象归元于云枢 | 深栈智启新纪元","2026-11-10 10:37:32",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add YYC-Cube\u002FYYC3-Knowledge-Base --skill agentic-jujutsu",{"tool":40,"command":41},"bunx","bunx skills add YYC-Cube\u002FYYC3-Knowledge-Base --skill agentic-jujutsu",{"tool":43,"command":44},"pnpm","pnpm dlx skills add YYC-Cube\u002FYYC3-Knowledge-Base --skill agentic-jujutsu","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Agent-智能体\u002FYYC3-Agent-智能体-ruflo\u002Fv2\u002F.claude\u002Fskills\u002Fagentic-jujutsu\u002FSKILL.md",true,[],[49,56,63,70,77,84],{"skillNo":50,"name":51,"nameZh":6,"description":52,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":19},"01M7CTDWHERCXPQCP0K03A86AS","vercel-sandbox","Run agent-browser + Chrome inside Vercel Sandbox microVMs for browser automation from any Vercel-deployed app. Use when the user needs browser automation in a Vercel app (Next.js, SvelteKit, Nuxt, Remix, Astro, etc.), wants to run headless Chrome without binary size limits, needs persistent browser sessions across commands, or wants ephemeral isolated browser environments. Triggers include \"Vercel Sandbox browser\", \"microVM Chrome\", \"agent-browser in sandbox\", \"browser automation on Vercel\", or any task requiring Chrome in a Vercel Sandbox.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fvercel-sandbox\u002FSKILL.md",[],[],{"skillNo":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":19},"01M7CTD8D2BNJY54S65M47HEJA","slack","Interact with Slack workspaces using browser automation. Use when the user needs to check unread channels, navigate Slack, send messages, extract data, find information, search conversations, or automate any Slack task. Triggers include \"check my Slack\", \"what channels have unreads\", \"send a message to\", \"search Slack for\", \"extract from Slack\", \"find who said\", or any task requiring programmatic Slack interaction.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fslack\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M7CTCSSB8ZP3Q58C3YZDZK1W","electron","Automate Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify, etc.) using agent-browser via Chrome DevTools Protocol. Use when the user needs to interact with an Electron app, automate a desktop app, connect to a running app, control a native app, or test an Electron application. Triggers include \"automate Slack app\", \"control VS Code\", \"interact with Discord app\", \"test this Electron app\", \"connect to desktop app\", or any task requiring automation of a native Electron application.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Felectron\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7CTCBJY76688V4DCZGWFHGQ","dogfood","Systematically explore and test a web application to find bugs, UX issues, and other problems. Use when asked to \"dogfood\", \"QA\", \"exploratory test\", \"find issues\", \"bug hunt\", \"test this app\u002Fsite\u002Fplatform\", or review the quality of a web application. Produces a structured report with full reproduction evidence -- step-by-step screenshots, repro videos, and detailed repro steps for every issue -- so findings can be handed directly to the responsible teams.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fdogfood\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7CTC4H0FFGVEAQR9PA4GA1A","agent-browser","Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to \"open a website\", \"fill out a form\", \"click a button\", \"take a screenshot\", \"scrape data from a page\", \"test this web app\", \"login to a site\", \"automate browser actions\", or any task requiring programmatic web interaction.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fagent-browser\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7CTBKVFYBG78FDSABSDJTTQ","shadcn","Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI, including chat interfaces. Provides project context, component docs, and usage examples. Applies when working with shadcn\u002Fui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for \"shadcn init\", \"create an app with --preset\", \"switch to --preset\", and for building a design system or applying a DESIGN.md or brand spec to shadcn\u002Fui.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-UI-界面组件\u002FYYC3-UI-界面组件-ui\u002Fskills\u002Fshadcn\u002FSKILL.md",[],[]]